GLM 4.6 needs about 219.2 GB of memory at Q4_K_M (a 215.6 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 27 quants, from 84.1 GB (TQ1_0) to 713.7 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 12.5 GB.
GLM‑4.6, a 360 B parameter mixture‑of‑experts model with a 200K‑token context window
GLM‑4.6 is a 360 B parameter model built on the glm4moe architecture. It supports a 200K‑token context window and is optimized for reasoning, coding, and agentic tasks. The model is released under the MIT license and is available in multiple GGUF quantizations.
Good at
- Long‑context reasoning
- High‑quality coding output
- Tool‑aware agentic behavior
Limits
- Requires large VRAM for full precision
- No vision support
- No explicit language support list
Best for
- Complex agentic workflows
- Large‑scale code generation
- Long‑form content creation
Who it’s for. Developers and researchers who need a large‑parameter model with extended context and tool‑use capabilities. The model is suitable for building agents that require reasoning and coding skills.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| TQ1_0 | 84.1 GB | 87.8 GB | Too big | Get file → | |
| IQ1_S | 96.9 GB (2 parts) | 100.6 GB | Too big | Get file → | |
| IQ1_M | 107.4 GB (3 parts) | 111.0 GB | Too big | Get file → | |
| IQ2_XXS | 115.4 GB (3 parts) | 119.0 GB | Too big | Get file → | |
| IQ2_M | 121.9 GB (3 parts) | 125.6 GB | Too big | Get file → | |
| Q2_K | 130.7 GB (3 parts) | 134.3 GB | Too big | Get file → | |
| Q2_K_L | 130.8 GB (3 parts) | 134.5 GB | Too big | Get file → | |
| Q2_K_XL | 134.7 GB (3 parts) | 138.3 GB | Too big | Get file → | |
| IQ3_XXS | 144.7 GB (3 parts) | 148.3 GB | Too big | Get file → | |
| Q3_K_S | 154.4 GB (4 parts) | 158.0 GB | Too big | Get file → | |
| Q3_K_XL | 158.2 GB (4 parts) | 161.9 GB | Too big | Get file → | |
| Q3_K_M | 170.6 GB (4 parts) | 174.2 GB | Too big | Get file → | |
| IQ4_XS | 190.7 GB (4 parts) | 194.3 GB | Too big | Get file → | |
| IQ4_NL | 201.6 GB (5 parts) | 205.2 GB | Too big | Get file → | |
| Q4_0 | 202.2 GB (5 parts) | 205.8 GB | Too big | Get file → | |
| Q4_K_S | 202.8 GB (5 parts) | 206.4 GB | Too big | Get file → | |
| Q4_K_XL | 203.7 GB (5 parts) | 207.3 GB | Too big | Get file → | |
| Q4_K_M | 215.6 GB (5 parts) | 219.2 GB | Too big | Get file → | |
| Q4_1 | 223.6 GB (5 parts) | 227.3 GB | Too big | Get file → | |
| Q5_K_S | 245.9 GB (5 parts) | 249.6 GB | Too big | Get file → | |
| Q5_K_XL | 252.1 GB (6 parts) | 255.7 GB | Too big | Get file → | |
| Q5_K_M | 253.2 GB (6 parts) | 256.8 GB | Too big | Get file → | |
| Q6_K | 293.1 GB (6 parts) | 296.7 GB | Too big | Get file → | |
| Q6_K_XL | 299.0 GB (6 parts) | 302.6 GB | Too big | Get file → | |
| Q8_0 | 379.3 GB (8 parts) | 382.9 GB | Too big | Get file → | |
| Q8_K_XL | 389.8 GB (8 parts) | 393.4 GB | Too big | Get file → | |
| BF16 | 713.7 GB (15 parts) | 717.4 GB | Too big | Get file → |
For 8 GB VRAM, use Q4_0 or Q4_K_S (≈205 GB file, ~206 GB memory at 8K context). For 16 GB VRAM, Q4_K_M or Q5_K_S (≈219–256 GB memory). For 32 GB VRAM, Q6_K or Q6_K_XL (≈296–303 GB memory). File sizes and memory estimates are approximate.
What GPU do you need for GLM 4.6?
These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: no quant fits at an 8K context.
- 6 GB VRAM: no quant fits at an 8K context.
- 8 GB VRAM: no quant fits at an 8K context.
- 12 GB VRAM: no quant fits at an 8K context.
- 16 GB VRAM: no quant fits at an 8K context.
- 24 GB VRAM: no quant fits at an 8K context.
- 32 GB VRAM: no quant fits at an 8K context.
- 48 GB VRAM: no quant fits at an 8K context.
Which GLM 4.6 quant should you choose?
Q4_K_M is the usual starting point for GLM 4.6: it balances file size and quality well. Here it is a 215.6 GB file.
The Q8_0 file is 379.3 GB, about 1.8 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.
If memory is tight, the smallest file is TQ1_0 at 84.1 GB, with the quality trade-offs that come with fewer bits. Lower bit counts mean smaller files and more quality loss.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about GLM 4.6
How much VRAM does GLM 4.6 need?
About 219.2 GB at Q4_K_M with an 8K context window. That is the 215.6 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run GLM 4.6 on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 87.8 GB, so you would need a larger GPU or CPU offloading.
What is the file size of GLM 4.6 GGUF?
It depends on the quant: TQ1_0 is 84.1 GB, IQ1_S is 96.9 GB, IQ1_M is 107.4 GB, IQ2_XXS is 115.4 GB, IQ2_M is 121.9 GB, Q2_K is 130.7 GB and 21 more.
What context length does GLM 4.6 support?
The GGUF header lists a trained context length of 202,752 tokens (198K). The KV cache grows with the context you use: about 3.1 GB at 8K and 12.5 GB at 32K.
Where can I download GLM 4.6 GGUF?
The files are on Hugging Face in the unsloth/GLM-4.6-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does GLM 4.6 use?
Hugging Face lists the license as mit. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run GLM‑4.6 on 8 GB VRAM?
Use a Q4_0 or Q4_K_S quantization; memory usage will be around 206 GB at 8K context, which exceeds 8 GB VRAM, so it is not feasible without further sharding.
Does GLM‑4.6 support vision tasks?
No, the model does not include vision capabilities.